January 1995, Volume 45, Issue 1

Original Article

Surgery for Bleeding Esophageal Varices

Mohammad Arshad Cheema  ( Department of Surgery, King Saud University, College of Medicine, Abha, Saudi Arabia. )
Mohammad Yahyà Ali Shehri  ( Department of Surgery, King Saud University, College of Medicine, Abha, Saudi Arabia. )


A total of 72 patients were operated for bleeding esophageal varices over five years. Cause of portai hypertension was cirthosis in 33, Schistosomal fibrosis in 23 and a combination of the two diseases in 3 cases. Biopsy was not available in 13 patients. Fifty-eight patients were child grade A and B, while 14 patients were grade C. Overall, there were 16 hospital deaths (22.2%) and 28 patients had complications (38.8%). Specifically, Hassab’s operation was done in 40 patients with 12.5% mortality and 11.7% incidence of rebleeding. Hassab’s operation plus esophageal transection in 13 patients was associated with 46.1% mortality and no incidence of rebleeding. Warren’s splenorenal shunt, done in 11 patients, was accompa­nied by 1 (9%) death and no incidence of rebleeding. Mortality rate increased significantly when esophageal transection was added to Hassab’s operation. It is concluded that for low risk patients being operated electively, Warren’s shunt is an acceptable alternative to Hassab’s operation which is better suited to emergency situations. Esophageal transection should not be added to Hassab\\\'s operation because this increases the mortality (JPMA 45:6,1995).


Portal hypertension (PH) is common in Saudi Arabia because of the high incidence of hepatitis1 and Schistosonil­asis2 Bleeding from esophageal varices (BEV) is a dreaded complication of PH. Operations for BEV have been associated with mortality as high as 60%3 and morbidity around 40%4. Studies from the Western hemisphere suggest that scierotherapy should be the initial management of these patients5,6. Nonetheless, about half of these patients will rebleed on sclerotherapy5,6 and will require ‘rescue surgery’ to control their bleeding. Moreover, sclerotherapy has not been successful in controlling fundal varices. There is no general agreement on the type of surgery to be performed on these patients. Basically operations to stop bleeding and prevent rebleeding in PH patients are either designed to devascularise the lower esophagus and stomach or they attempt to shunt blood away from the high pressure portal circulation to systemc veins. Because of considerable difference in pa­tients’ populations with varying etiologies of PH in different parts of the world, care has to be taken to translate the results of treatment for BEV from one society to the other7. The purpose of this study is to present the experience of surgical management and outcome for bleeding esophageal varices patients inAsir Central Hospital (a tertiary care facility affiliated with College of Medicine Abha. Saudi Arabia).

Patients and Methods

All patients aged above 12 years operated forBEV from August, 1988 to October, 1993 were included. Data was especially collected regarding the type of liver disease as well as Child’s grade. The patients were divided into three main groups, which were Hassab’s operation, Hassab’s operation combined with esophageal transection and Warren’s shunt. The patients in the three main groups were comparable with regards to age and sex and the type of liver disease (Table I).

Eight patients who did not fall in the above three groups were grouped as a miscellaneous group. Hassab’s operation involved devascularisation of lower esophagus and upper stomach and splenectomy as described by Hassab8. Hassab’s operation and esophageal transection (ET) was a transabdominal devascularisation similar to the above with addition of transection and anastomosis of esophagus using EEA stapler. Warren’s shunt consisted of anastamosis of distal splenic vein to left renal vein as described by Warren et al9.  No attempt was made L achieve a complete splenopancreatic disconnection. The proportion of deaths for the three types of opera­tions were analyzedby using Chi-square, Z-test of proportions and Fisher’s Exact tests at 5% level of significance.


Seventy-two patients were operated for BEV. This included 61 males and 11 females with a median age of 39 years (range 14-70 years). All patients were treated initially with sclerotherapy and surgery performed when this therapy failed or patient was noted to have fundal varices. Thirty patients had Child’s grade A, while 28 and 14 were grade B and C respectively. Histopathology of liver showed cirrhosis in 33 and Schistosomiasis in 23 patients. Three further patients had combined cirrhosis and Schistosomiasis and fourpatients with cirrhosis had superadded hepatocellular carcinoma. Unfortu­nately 10 patients from Hassab’s operation group and three patients from Hassab +OT group did not have liverbiopsy. No patient volunteered histoiy of alcohol intake. Forty patients had Hassab’s operation alone, 13 had Hassab’s and ET and 11 had Warren’s shunt. Of the remaining patients, 3 had under-running ofvarices, two had unsuccessful attempts at portacaval shunt, one each had splenectomy alone and esophageal transection alone. One patient died on table from bleeding before any operation: Only one patient from Warren’s shunt group had emergency operation, while 66% patients in the other two main groups had emergency procedures. Deaths and important complications following the three common operations are outlined in Table II.

Mortality was significantly increased (Z=2.28, P<0.05) when esophageal transection was added to Hassab’s operation (Table III).

Of the six patients who died following Hassab and ET. four died from continued bleeding, one from a confirmed esophageal leak and one developed multiple problems which included sepsis from suspected esophageal leak. Overall 28 patients (38.8%) had complications. Commonest complication was failure of the operation to stop bleeding resulting in death of nine patients in which it occurred. Renal failure in 3 patients was also fatal. Sixteen patients (22.2%) died following operations within 30 days of admission (Table IV).

Mortality in 58 patients with low risk Child grade A and B disease was 15.5% and rose to 50% in high risk patients with Child grade C liver disease (X2=7.82, P


Surgical management of BEV remains both frustrating and challenging. An operation for BEV has to be judged from triple aspects of mortality, incidence of rebleeding and hepatic encephalopathy. Although Hassab’s operation and Warren’s shunt had comparable mortality in the low risk Child A and B patients (3/34, (8.8%) and 1/11 (9.0%) respectively). When esophageal transection was added to Hassab’s operation, mortality even in low risk patients was prohibitive (4/9, 44.4%). Esophageal transection is known to be associated with high mortality4. Although Suigura et al10 and Johnston11 have reported 5.2% and 14% mortality following esophageal transection respectively, it has ranged from 35% to 60% in other reports12-15. Transection of esophagus following sclerotherapy has a high incidence of leakage from esopha­gus15 and as noted by Zhang16, in a controlled prospective comparison of Hassab’s operation with and without esopha­geal transection, it increases early morbidity without any long term benefit Rebleeding (4/34= 11.7%) was only noted following Hassab’s operation. Two of these patients were successfully treated by sclerotherapy. Sakai et al17 showed that scierotherapy was effective in the control of rebleeding in 97.3% of the patients in the group who had previous portal hypertension surgery and in 72.7% of those without previous portal hypertension surgery. Encephalopathy occurred mainly in Child C patients following devascularisation procedures. Our initial experi­ence with Warren’s shunt is encouraging. No patients devel­oped encephalopathy following Warren’s shunt. On the other hand, Stauss et al18 from Brazil have reported 14.8%patients, developing encephalopathy after Warren’s shunt. Ezzat et al19 reported clinical encephalopathy in 15% patients after War­ren’s shunt but when they looked at Schistosomial portal fibrosis subgroup of patients, the incidence was only 4.4%. The short followup period may have been the reason for the low incidence of encephalopathy in our patients since Ezzat et al reported the mean duration for onset of encephalopathy as 32±23 months. Warren’s shunt is a technically demanding procedure unsuitable for use in an emergency situatioa In contrast, Hassab’s operation can be done with less difficulty in an emergency. Based on the findings from this study, it is recommended that for a relatively low risk patient being operated electively, given the expertise, Warren’s shunt is an appropriate operation because of acceptable morbidity and mortality. If the same patient is being operated as an emergency, Hassab’s operation should be chosen because of acceptable efficacy and relative ease of performance. Special attention should be paid to correct, if possible any disturbed coagulation parameters. An ideal operation for a poor risk patient, who is bleeding inspite of scierotherapy, has yet to be found. Addition of esophageal transection increases the mortality significantly and although there is an increased incidence of rebleeding if the esophagus is not transected, scierotherapy may be useful in this situation as we found in two of our patients. Recognition of the need for further study of this problem has led to the formation of Asir Portal Hypertension Study group and Asir Portal Hyperten­sion registry.


1. Al-Faleh, F.Z. Hepatitis B infection in Saudi Arabia. Ann. Saudi Med., 1 988;8:474-80.
2  Islam, S.S. Schistosomiasis control in Saudi Arabia. Dammam, Proceedings of the Fourth Saudi Medical Conference, 1978;pp.63- 71.
3. Terblanche, .T. The surgeon’s role in the management of portal hypertension. Ann. Surg., 1989;209:381-95.
4. Keagy, BA., Schwartz, J.A. and Johnson, 0. Should ablative operations beused for bleeding esophageal varices? Ann. Surg., I 986;203:463-69.
5. Warren, W.D., Henderson, J.M, Milikan, W.J. et al. Distal splenorenal shunt versus endoscopic sclerotherapy for long term management of variceal bleeding. Preliminary report ofa prospective randomized trial. Ann. Surg., 1986;203 454-62.
6. Westaby, D., Mcdougall, B.R.D. and Williams. R. Improved survival following injection sclerotherapy for bleeding varices: final analysis of a controlled trail. Hepatology, 1985;5:827-30.
7. Johnson, G.W. Esophageal transaction. In surgery of the liver andbiliary tract. Ed. Blumgart LH. Edinburgh, Churchill Livingstone, 1988;pp. 1369-78.
8. Hassab, MA. Nonshunt operations in portal hypertension without cirrhosis. Surg. Gynsecol. Obstet., 1970;131 :648-54.
9. Warren, W.D.. Zeppa, R. andFomon, J.J. Selective trans-splenic decompression of gastroesophageal varices by distal splenorenal shunt. Ann. Surg., 1967;166:437-55.
10. Suigura, M. and Futagawa, S. Results of six hundred thirty six esophageal transactions with paraesophageal devascularisations in the treatment of esopha­gealvarices. J. Vase. Surg., 1984;1:254..60.
11. Johnston, G.W. Six years experience of esophageal transaction for esophageal varices using a circular stapling gun. Gut, I 982;23:770-73.
12. Burroughs, AK., Hamilton, G., Phillips, A. et al. A comparison of sclerotherapy with staple transaction of the esophagus for the emergency control of bleeding from esophageal varices. N. Engl. J. Med., I 989;32 1:857-62.
13. Gouge, T.H. and Ranson, J.H.C. Esophageal transaction with paraesophageal devascularisation (the Suigura procedure) for bleeding esophageal varices. Am. J. Surg., 1986;151:47-53.
14. Wanamaker, S.R., Cooperman, M and Carey, L.C. Use of EEA stapling instrument for control ofbleeding esophagealvarices. Surgery. 1983;94:620-26.
15. Griffin, W.O., Keagy, B.A., Schwartz, J.A. et al. Should ablative operations be used for bleeding esophageal varices? Ann. Surg.. 1 986;203 :463-69.
16. Zhang. DC. Hassab’s procedure with or without lower esophageal transaction in the treatment of portal hypertension. A prospective controlled study [translated from Chinese]. Chung. Hua. Wai. Ko. Tsa. Chih., 1991;29:561-3,589-90.
17. Sakai, P., Boaventura, S., Ishioka, S. etal. Sclerotherapy of bleeding esophageal varices in Schistosomiasis. Comparative study in patients with and without previous surgery for portal hypertension. Endoscopy, 1 990;22:5-9.
18. Strauss, E., Raia, S., Da Silva, L.C. at al. Portal hypertension in Schistosomiasis. Long term follow up of a randomized trial comparing three types of surgery. Hepatology, 1990;12:854 [Abstract A-66].
19. Ezzat, F.A., Abu-elmagd, KM., Sultan, A.A. et al. Schistosomial versus Non-Schistosomial varicial bleeders:Do they respond differently to selective shunt?Ann. Surg., 1989;209:489-99.

Journal of the Pakistan Medical Association has agreed to receive and publish manuscripts in accordance with the principles of the following committees: